You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@cassandra.apache.org by folex <0x...@gmail.com> on 2015/09/14 13:12:11 UTC

cassandra-stress where query with greater than operator

Hi All.

I'm trying to set up cassandra load testing and came up with the next YAML
config (https://gist.github.com/folex/d297cc8208a2e54a36d7) :

keyspace: stress

keyspace_definition: |
  CREATE KEYSPACE stress WITH replication = {'class': 'SimpleStrategy',
'replication_factor': 3};

table: messages

table_definition: |
  CREATE TABLE stress.messages (
    chat_id bigint,
    uid_to bigint,
    message_id bigint,
    text blob,
    uid blob,
    PRIMARY KEY ((chat_id, uid_to), message_id)
  ) WITH CLUSTERING ORDER BY (message_id ASC)

columnspec:

  - name: text
    size: gaussian(0..44)

  - name: uid
    size: gaussian(0..32)

  - name: chat_id
    size: uniform(0..64)
    population: uniform(500..2K)

  - name: uid_to
    size: uniform(0..64)
    population: uniform(500..2K)

  - name: message_id
    size: uniform(0..64)

insert:
  partitions: fixed(1)
  batchtype:  UNLOGGED
  select:     fixed(1)/1

queries:
  scan:
    cql: select * from stress.messages where chat_id = ? and uid_to = ? and
message_id > ? limit 50;
    fields: samerow


Now, insert works as expectet, but "scan" just hangs up after printing out
stat headers.
Also I found that if I remove "and message_id > ?" part, scan start working.

Is that a known problem? How do I workaround that?

Thanks in advance.

- folex